Contracted revenue. Milestone-gated capital.
Thorium One Power develops behind-the-meter power and water infrastructure for the AI data center buildout: two mirrored 20-year contracts with a single campus offtaker, capital phased against licensing and construction milestones, and an advanced-nuclear layer structured as an upgrade to an already-contracted position — never a precondition of it.
The market thesis
The constraint on AI compute has moved from chips to powered land. Hyperscale operators can fund campuses faster than the grid can energize them: the median U.S. interconnection request took more than five years to reach commercial operation for projects built in 2025, and large loads that trigger transmission upgrades routinely face four-to-seven-year timelines. That gap is why bring-your-own-power development — independent generation co-located with load — has become a category rather than a workaround.
Capital has noticed the power half of that trade. What remains structurally under-supplied is the second constraint: water. A gigawatt campus on hybrid cooling consumes millions of gallons a day, and in the arid basins where cheap gas and open land actually exist, freshwater is the binding permitting and social-license risk. Developers selling megawatts alone leave that problem with the offtaker.
Revenue model
Revenue is contracted, not merchant. The campus sells into two mirrored long-term agreements with the same offtaker, both beginning at first power:
| Contract | Structure |
|---|---|
| Power purchase agreement | 20-year tenor with a capacity-payment floor and liquidated damages on commercial operation dates. Two tranches: a bridge tranche priced at signing, and a fixed clean-firm nuclear tranche that phases in as molten-salt modules come online. |
| Water service agreement | Take-or-pay for distillate-quality cooling water, with availability guarantees and commercial-operation-date damages mirroring the PPA. |
| Upside layers (not in base case) | Mineral recovery from the desalination concentrate stream (iodine, lithium options), grid export once interconnection completes, and campus expansion pads as offtaker load grows. |
The commercial design deliberately keeps the offtaker's obligations independent of the nuclear schedule: the bridge fleet alone is sized to serve the full contracted obligation, so the nuclear phase changes blended cost and carbon intensity rather than whether power is delivered.
How capital is phased
Risk is retired before capital commits, in that order — not the reverse.
Development capital
Carries site control, offtaker contracting, water-supply agreements, interconnection position, permitting and the licensing path. This is the stage that converts an idea into a contractable position.
Project capital — conventional infrastructure
The bridge gas fleet with storage and the thermal desalination plant are proven, financeable technology underwritten against long-term contracted revenue from an executed PPA and water service agreement. Nothing here depends on first-of-a-kind performance.
Nuclear capital — committed against evidence
Molten-salt module capital commits only after licenses are in hand, behind milestone gates, with multiple advanced-reactor paths preserved so no single vendor's schedule sits on the critical path. The nuclear layer improves an already-contracted asset.
Why the position is defensible
Water, not turbines
Gas turbines can be bought by anyone. Produced-water supply agreements, thermal desalination driven by the power island's own waste heat, and a zero-freshwater permitting story are what competing gas-only developers in West Texas do not have — and cannot assemble quickly.
One counterparty, two contracts
Selling power and water together raises contract value per campus and raises switching costs, because the offtaker is not assembling and managing separate power, water and interconnection counterparties.
A federal window for advanced nuclear
The ADVANCE Act and the DOE Reactor Pilot Program opened the most favorable U.S. licensing environment for advanced reactors in decades — the window the nuclear phase is designed to use, without betting delivery on it.
Treating water the basin must dispose of
The model takes produced water operators otherwise pay to inject and turns it into product, reducing deep-injection volume for the water treated rather than adding to it — aligning the project with the basin's own disposal and seismicity pressures.
Risk, stated plainly
An investment case that lists only strengths is not a diligence document. The material risks in this model are:
| Risk | Why it matters |
|---|---|
| Counterparty concentration | The largest single risk. A campus anchored by one or two offtakers means revenue depends on a small number of contracts, so counterparty credit, security packages and contract terms carry far more weight than in a diversified merchant portfolio. |
| Development-stage execution | The company is pre-FID. Site control, offtaker contracting, water supply and interconnection all have to be converted from position into executed agreements before project capital is underwritable. |
| Gas price and supply | The bridge tranche is served by gas. Fixed-price supply and firm transport are the intended mitigants, but basis, transport and supply performance remain real exposures during the bridge phase. |
| Water treatment performance | Produced-water chemistry varies. Delivering distillate on spec, continuously, at industrial scale is an operating discipline, and concentrate handling carries NORM and disposal obligations. |
| Construction and COD | Commercial-operation-date damages sit with the developer for both plants — that protects the offtaker and concentrates schedule risk on this side of the table. |
| Nuclear licensing and schedule | First-of-a-kind licensing risk is real. It is deliberately isolated so that slippage delays a price step-down rather than power delivery — but it remains a risk the sponsor carries. |
| Regulatory change | Large-load and co-location rules are actively evolving in ERCOT, including under Texas Senate Bill 6. Rule changes can affect campus configuration and economics. |
Stage
Thorium One Power is development-stage and pre-FID. The indicative schedule places capacity reservation and letters of intent in 2026, final investment decision and execution of the mirrored agreements in 2027, first power and first water in 2028–30, and the molten-salt nuclear phase-in from 2033. Leadership, development partners and capital sources are reviewed under NDA as part of the investor diligence package.
Questions capital partners ask
What is the investment case for Thorium One Power?
How is Thorium One Power financed?
Where does Thorium One Power's revenue come from?
What is the biggest risk in the Thorium One Power model?
Is Thorium One Power a bet on nuclear technology?
How do investors engage with Thorium One Power?
Request the diligence package
Market thesis, phased capital plan, site and water position, commercial structure and the nuclear phase-in schedule — shared under NDA with institutional investors, family offices and strategic partners.
Important: nothing on this website is an offer to sell, or a solicitation of an offer to buy, any security, and no such offer will be made except through definitive offering documents to eligible investors. Campus capacity, water and date figures are targeted and indicative, subject to diligence and final engineering. Interconnection timing figures are from Lawrence Berkeley National Laboratory's Queued Up 2026 Edition. Related reading: about the company · interconnection queue field guide · produced-water cooling field guide.